* Greg Folkert (greg@???) [030925 12:34]:
> --
> On Thu, 2003-09-25 at 14:48, George Szekely wrote:
> > Dear List:
> >
> > fakereject was the perfect solution for notifying a sender that .exe
> > is not acceptable (any file must be zipped).
> >
> > It only sent back a custom 550 ... without the attachement.
> >
> > I wonder why it was taken out of the latest release.
> >
> >
> > # warn log_message = DISCARD: ATTACHMENT ($found_extension)
> > # demime = scr:pif:exe:com
> > # control = fakereject
> > # discard demime = scr:pif:exe:com
>
> Do yo not understand that this exactly doubles the amount of mail
> associated with a piece of SPAM. (or W32Swen/Gibe@mm)...
Does a 5xx really do this? I understand that generating a bounce
message does, but as I understand it, a 5xx is really the way to go. In
this case, if the peer is the virus' dumb smtp engine, the 5xx just ends
the transaction. (I assume the virus isn't going to generate a bounce!)
If the peer is a relay, the problem is there, not on your server. They
should have issued a 5xx the first time they received the virus rather
than ever accepting it in the first place. I suppose this is an
idealized case -- in the real world, some virus and spam come from
relays, and a bounce will be generated (by the relay), which is bad.
Am I wrong in thinking that a 5xx at smtp time is a good way to go?
Maybe in the special case of a message which has been verified as a
virus carrier, a silent accept and drop is good, but I think in the
general case a 5xx is the right thing. Am I wrong?
good times,
Vineet
--
http://www.doorstop.net/
--
"As we enjoy great advantages from inventions of others, we should be glad
of an opportunity to serve others by any invention of ours; and this we
should do freely and generously." --Benjamin Franklin
